In Conversation: Costas Meghir and Amer Hasan on early childhood development

Robust evidence demonstrates that investments in early childhood development (ECD) are critical for long-term growth and development. ECD interventions have been shown to enhance children’s cognitive abilities, language acquisition, and socioemotional skills – with positive effects on long-run educational, employment, and health outcomes. Yet ECD systems in many low- and middle-income countries remain weak or underdeveloped.

Costas Meghir, the Douglas A. Warner III Professor of Economics and a faculty affiliate of the Economic Growth Center (EGC) and Yale Inclusion Economics (YIE), has designed and tested several ECD interventions – including an ongoing study in rural Kenya. Amer Hasan, Senior Economist with the World Bank’s Education Global Practice, has led research and policy engagement on several early learning projects, focusing on South Asia and East Asia.

This conversation took place on the sidelines of the “Designing quality care: Insights on advancing child development and women’s empowerment” conference in New Haven. Meghir and Hasan discussed lessons from their ECD interventions, remaining gaps in the research, and the importance of translating evidence into policy and practice at scale. We look forward to building on the themes explored in this discussion at our upcoming event in Nairobi on June 30, “Harnessing Human Capital for Growth and Development in Kenya.”

How does ECD foster growth and development?

Costas Meghir: Early childhood development lays the foundation for education and skill accumulation. This in turn helps people work more effectively, adopt innovations ­– generally and in agriculture specifically – and allows them to better engage with economic progress, improving their standards of living. An educated workforce helps create an attractive environment for investment, which is essential to growth.

Many advocates also argue that child development and education are foundations of freedom and human rights. Lack of literacy and basic education has significant political and social implications: basic skills are essential for engaging in the world, like participating in democracy.

Amer Hasan: Ninety percent of brain development takes place during the first five years of life. Across many settings, investing in the early years has been shown to be one of the smartest investments governments can make – especially in low- and middle-income countries.

How has the field of ECD research evolved in recent years?

Hasan: Early on, there were the Perry Preschool and Abecedarian studies, which were both US-centric, and the long-term Jamaica study. In the last 15 years, there's been a tremendous proliferation of evidence from low- and middle-income countries showing the importance of early-year investments for children, parents, and communities. Disciplines have come together; economists and child development specialists are increasingly working together – deepening our understanding of ECD using an array of multidisciplinary tools and approaches.

Meghir: We’ve moved beyond asking whether providing a stronger developmental and educational foundation is effective for children’s development. The emphasis now is on understanding what will work at scale. What are the best ways to deliver interventions? Should they be complemented with health interventions? How can we make the case to budget-constrained governments?

What are the key indicators for measuring ECD?

Hasan: ECD is often shorthand for a set of outcomes, including health, motor skills, cognitive development, and socio-emotional skills – all of which are key building blocks for a productive life. The World Bank, along with its partners, has actively participated in the development of free, standardized measurement tools (such as AIM-ECD) as well as classroom observation tools to understand how teachers interact with kids (such as TEACH and TEACH ECE).  

Meghir: Ultimately, we want to see improved standards of living through increased wages and employment and improved health. But such follow-ups naturally take time, and we can't wait that long, particularly since some long-term studies have already provided proof of concept. So, as Amer says, we look at cognitive and socio-emotional development in the short- and medium-term, a few years after the intervention.

Costas, is there an ECD intervention where the outcomes surprised you?

Meghir: Our large NIH-funded project in the Indian state of Odisha looked at the potential for delivering an ECD intervention in a group setting. In the standard Reach Up program, home visitors make individual household visits, leading mothers and children through activities on language, cognition, and motor skills. It works well, but the one-on-one approach is quite expensive. We expected group delivery to be cheaper, but with weaker effects. However, we were pleasantly surprised to see impacts as large as – or even larger, if we allow for differences in attendance – than the standard approach. 

We realized we were fostering connections among socially isolated Indian women – who often marry into different villages where they don’t have much social interaction. Group delivery creates a platform for that; they become friends. It can also be culturally transformational: interacting with young children isn’t universally practiced, but in groups, they see other mothers doing it. At scale, the group model would cost at most $37 per child per year – about one-fourth the cost of home visiting. Our approach has drawn a lot of attention and further research.

Amer, your World Bank research is often in partnership with governments. How has ECD evidence translated into policy?

Hasan: In Indonesia, we evaluated a government effort to expand access to community-based preschools (“playgroups”) to about 3,000 villages. Villages received ECD orientation and block grants of about $18,000 over three years to establish centers for children aged 0 to 6. Community volunteers served as teachers, who received about 200 hours of training.

Following cohorts of one-year-olds and four-year-olds from 2009 to 2016, we compared outcomes for both cohorts at age 8. We found stronger outcomes among the younger cohort, which was 23 percentage points more likely to have ever enrolled in playgroups than the older cohort and also attended these playgroups for five months longer. There is evidence that these impacts lasted beyond the project’s duration, as well.

More broadly, the intervention helped provide actionable evidence for a government seeking to implement an ambitious set of reforms for the development and expansion of holistic, integrated ECD.

Costas, what’s distinctive about your ongoing ECD study in Kenya?

Meghir: This preschool intervention has two aims: to improve child development and create employment opportunities for women. Babysitting alone, where women drop children off while working, can deprive children of stimulation and nurturing and is a lost opportunity. Our intervention creates preschool positions for kids, starting at age three, in a stimulating environment based on ECD principles. We added teachers, provided on-the-job training and materials (toys, books, puzzles), and created activity guides for the teachers. The outcomes are spectacular: a 0.5 standard deviation improvement in IQ – indicating broader impacts than merely testing kids on what they’re taught – and large improvements in women’s earnings, particularly among highly income-constrained groups.

We are also seeing increased school participation for older siblings, who had likely been caring for younger siblings while their mothers worked. These policies can reshape household dynamics: mothers gain employment opportunities and income autonomy, children get better educational outcomes, and siblings return to school.

However, this is not a panacea ­– the region remains extremely poor. In addition to ECD, the government should also invest in other growth drivers.

What are the key gaps in ECD research and policy?

Meghir: Aggregating experience from a broad body of work (my own and others), I would say we lack a good understanding of how ECD impacts vary by level of deprivation. Experimental work in developing countries seems to suggest they are most effective in contexts of high malnutrition, while they seem to work less well for the non-malnourished poor. We need an improved understanding of how to implement and target interventions, so they can improve outcomes for all children in deprivation, not only the worst off of those.

Another gap is understanding why many nutritional interventions have had fairly limited effects on cognitive development, despite strong correlations between malnutrition and performance. In one Colombia-based study, we showed that starting a nutrition and stimulation intervention very early could improve height, as well as cognition. We also need to better understand what makes interventions sustainable. Some effects last, like in Jamaica and ours in Cuttack, while others fade.  

All this should not delay widespread implementation: we know enough to start implementing and we will learn more as the interventions scale. Ongoing research remains important, but we have enough evidence to advocate for putting high-quality, well-targeted ECD programs in place.

Hasan: To this list of gaps, I would add that we often don’t know enough about the cost-effectiveness of various interventions. Fortunately, that’s starting to change. One example is the Global Education Evidence Advisory Panel, co-hosted by the World Bank and several development partners, which recently released guidance on “smart buys” and “bad buys” in education, based on the evidence.

More broadly, these issues often need multi-layered solutions. In Pakistan, discussions around stunting now focus on enteric disease as well as nutrition – especially in rural areas, where the absence of water and sanitation facilities plays a big role. In such contexts, nutrition-only interventions may be less effective than combining them with waste management or water and sanitation. Combining approaches is useful for policymakers, who often have limited time and want to try multiple approaches. Recently, for example, researchers are combining multiple “smart buys” – and they’re finding very large effects in certain settings.

Meghir: Packaging interventions is also critical. As researchers, we like testing a single margin. But in practice and in policy design, we should be interested in complementary approaches. We cannot run infinite experiments. We need to combine our current empirical knowledge with theoretical reasoning of how to combine what works.